Commit Graph

10066 Commits

Author SHA1 Message Date
Dominik Dabek
e6460e5534 Add setArgBuffer caching to l0
When trying to set the same values we don't need to set it again.
The pointer and allocId need to be the same.

Related-To: NEO-6737

Signed-off-by: Dominik Dabek <dominik.dabek@intel.com>
2022-03-11 14:20:15 +01:00
Filip Hazubski
3eab7009ac Move SCM related WAs logic from CSR to EncodeComputeMode
This will help with unifying the logic between APIs and GENs.

Related-To: NEO-6728

Signed-off-by: Filip Hazubski <filip.hazubski@intel.com>
2022-03-11 14:00:53 +01:00
Kamil Kopryk
7d6bee26c7 Move pvc helpers to pvc files
Related-To: NEO-6631
Signed-off-by: Kamil Kopryk <kamil.kopryk@intel.com>
2022-03-11 13:46:59 +01:00
Rafal Maziejuk
3490b489ad Correct IMAGE1D_BUFFER width size calculation in BCS
Buffer's default bytesPerPixel value always equals 1 and as
IMAGE1D_BUFFER is originally an image, X coordinate needs to be
multiplied by bytesPerPixel in both copySize and (src/dst)Size.

Signed-off-by: Rafal Maziejuk <rafal.maziejuk@intel.com>
Related-To: NEO-6134
2022-03-11 09:34:40 +01:00
Michal Mrozek
6b29b03c29 Change default thread arbitration policy on PVC.
New default is round robin after stall.
Resolves: NEO-6731

Signed-off-by: Michal Mrozek <michal.mrozek@intel.com>
2022-03-11 09:20:53 +01:00
Joshua Santosh Ranjan
c7ce397b17 Metrics IP Sampling Add Support for Streamer APIs
This Patch adds support for collecting IP Metrics using
StreamerOpen, StreamerClose and StreamerReadData

Related-To: LOCI-2755
Related-To: LOCI-2756

Signed-off-by: Joshua Santosh Ranjan <joshua.santosh.ranjan@intel.com>
2022-03-11 03:03:22 +01:00
Brandon Yates
16b0ebe372 Update manifest to L0 loader 1.7.15
Signed-off-by: Brandon Yates <brandon.yates@intel.com>
2022-03-11 02:20:53 +01:00
Grzegorz Choinski
5183de21fa fix finding metrics when found empty
Do not try to include empty directories, as cmake will not
explicitely return directiories, which would be searched
by compiler anyway. Such found should still be considered
as successful.

Signed-off-by: Grzegorz Choinski <grzegorz.choinski@intel.com>
2022-03-10 23:32:33 +01:00
Bartosz Dunajski
c6e27bdc18 Enable prepatcher DirectSubmission WA
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2022-03-10 20:23:11 +01:00
Bartosz Dunajski
79fedd59e4 Add BB_START prepatching WA - disabled by default
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2022-03-10 17:05:53 +01:00
Krystian Chmielewski
1b4319fdab Update zebin note target metadata
This commit removes ZebinTargetMetadata struct, and uses
ZebinTargetFlags for both target validations: via machine type, and
via intel gt notes.

Signed-off-by: Krystian Chmielewski <krystian.chmielewski@intel.com>
2022-03-10 16:47:01 +01:00
Bartosz Dunajski
b8028d79c7 PVC: Fix compute units for scratch calculation
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2022-03-10 14:04:58 +01:00
Lukasz Jobczyk
c8ba97e492 Restore gpgpu csr's mutex lock in the enqueue blit
Signed-off-by: Lukasz Jobczyk <lukasz.jobczyk@intel.com>
2022-03-10 13:36:46 +01:00
Bartosz Dunajski
b8d5fac10f Add missing lock in MapOperationsHandler
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2022-03-10 13:17:46 +01:00
Bartosz Dunajski
e24322f266 Debug flag to control MI_ARB_CHECK prefetcher
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2022-03-10 12:50:05 +01:00
Compute-Runtime-Validation
41c51df086 Revert "Recycle old command buffers of immediate command lists"
This reverts commit 8f93f4f3ec.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2022-03-10 09:12:47 +01:00
ocldev
fe53879f16 dependencies update
Signed-off-by: ocldev <ocldev@intel.com>
2022-03-09 22:20:33 +01:00
Compute-Runtime-Validation
1b1481e587 Revert "Add report of extension cl_intel_split_work_group_barrier"
This reverts commit 886d4a36f7.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2022-03-09 20:11:27 +01:00
Grzegorz Choinski
6133d8c4b5 fix paths to prebuilt kernels
It otherwise looks for different path to copy kernel than it is generated into

Signed-off-by: Grzegorz Choinski <grzegorz.choinski@intel.com>
2022-03-09 18:55:44 +01:00
Grzegorz Choinski
ff9522f8ea infra update
Signed-off-by: Grzegorz Choinski <grzegorz.choinski@intel.com>
2022-03-09 18:18:21 +01:00
Patryk Wrobel
835b344968 Add debug flag to disable GPU hang detection
This change introduces the new flag called DisableGpuHangDetection.
By default it is disabled. When someone wants to disable hang checking,
then this flag can be set to true.

Related-To: NEO-6681
Signed-off-by: Patryk Wrobel <patryk.wrobel@intel.com>
2022-03-09 18:07:42 +01:00
Lukasz Wesierski
9d4dacacca Add report of extensions
cl_intel_subgroup_matrix_multiply_accumulate and
cl_intel_subgroup_split_matrix_multiply_accumulate

Related-To: NEO-6745
Signed-off-by: Lukasz Wesierski <lukasz.wesierski@intel.com>
2022-03-09 17:47:42 +01:00
Lukasz Wesierski
886d4a36f7 Add report of extension cl_intel_split_work_group_barrier
Related-To: NEO-6746
Signed-off-by: Lukasz Wesierski <lukasz.wesierski@intel.com>
2022-03-09 17:40:21 +01:00
Milczarek, Slawomir
c0b7f05897 Add memory prefetch for kmd migrated shared allocations
This feature is disabled by default, controlled with the knob
AppendMemoryPrefetchForKmdMigratedSharedAllocations

Related-To: NEO-6740

Signed-off-by: Milczarek, Slawomir <slawomir.milczarek@intel.com>
22.11.22682
2022-03-09 16:02:18 +01:00
Daniel Chabrowski
10e7b9d5be Test prelim buffer object ext
Signed-off-by: Daniel Chabrowski daniel.chabrowski@intel.com
Related-To: NEO-6591
2022-03-09 12:39:43 +01:00
Mateusz Jablonski
3e8a6680e2 Allow undefined symbols in symbol table for OCL
Related-To: NEO-5081
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
2022-03-09 10:32:15 +01:00
Kamil Kopryk
038d1d54fa Correct xe_hpc tests
Related-To: NEO-6631


Signed-off-by: Kamil Kopryk <kamil.kopryk@intel.com>
2022-03-09 09:21:30 +01:00
Filip Hazubski
43ed1acc63 Update StateComputeModeProperties design
Related-To: NEO-6728

Signed-off-by: Filip Hazubski <filip.hazubski@intel.com>
2022-03-08 16:30:25 +01:00
Michal Mrozek
cd15c82eab Do not prefer copy engine for local to local transfers.
Execution Units are faster.
Signed-off-by: Michal Mrozek <michal.mrozek@intel.com>
2022-03-08 15:42:58 +01:00
Zbigniew Zdanowicz
8f93f4f3ec Recycle old command buffers of immediate command lists
Related-To: NEO-6242

Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2022-03-08 14:59:26 +01:00
Filip Hazubski
dd01cff879 Unify logic determining thread arbitration policy value
Related-To: NEO-6728

Signed-off-by: Filip Hazubski <filip.hazubski@intel.com>
2022-03-08 13:14:56 +01:00
Dominik Dabek
d5fedf90c5 Fix for svm pointer arg caching
Previous version could cause segfaults.

Related-To: NEO-6737

Signed-off-by: Dominik Dabek <dominik.dabek@intel.com>
2022-03-08 12:13:15 +01:00
Filip Hazubski
865ef4d039 Add RPL-S device IDs
Signed-off-by: Filip Hazubski <filip.hazubski@intel.com>
2022-03-08 11:23:57 +01:00
Jaime Arteaga
e959597546 Dont return on void function
Signed-off-by: Jaime Arteaga <jaime.a.arteaga.molina@intel.com>
2022-03-07 21:48:38 +01:00
Lisanna Dettwyler
e6e6ae2d2b Don't use low-priority CSR for copy-only cmdq
getCsrForLowPriority will always return a CSR with a compute engine
regardless of the ordinal used. When a copy-only ordinal is used to
create a command queue or command list, batch buffers are written using
BCS style commands rather than CCS style. This caused BCS commands to be
submitted to a low-priority compute engine which was invalid. This patch
ensures a copy engine is selected when a copy-only ordinal is used even
if ZE_COMMAND_QUEUE_PRIORITY_PRIORITY_LOW is specified.

Related-To: LOCI-2980

Signed-off-by: Lisanna Dettwyler <lisanna.dettwyler@intel.com>
2022-03-07 21:38:12 +01:00
Zbigniew Zdanowicz
8f5dd3cff5 Refactor immediate command list flushing
Related-To: NEO-6242

Signed-off-by: Zbigniew Zdanowicz <zbigniew.zdanowicz@intel.com>
2022-03-07 20:37:28 +01:00
Filip Hazubski
80b520bc9b Change ThreadArbitrationPolicy enum type to int32_t
Change ThreadArbitrationPolicy::NotPresent value to -1
Update initial values to ThreadArbitrationPolicy::NotPresent

Related-To: NEO-6728

Signed-off-by: Filip Hazubski <filip.hazubski@intel.com>
2022-03-07 20:04:24 +01:00
Bartosz Dunajski
c612480f16 Add DG2 device ids
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
2022-03-07 19:53:28 +01:00
Daniel Chabrowski
de0361cfa5 Test prelim VM bind, debug
Signed-off-by: Daniel Chabrowski daniel.chabrowski@intel.com
Related-To: NEO-6591
2022-03-07 19:42:41 +01:00
Krystian Chmielewski
439389ae19 Do not resolve external functions when none passed
This commit adds check in Linker::resolveExternalFunctions checking
if external functions are present before trying to resolve dependencies
and adds default values for ExternalFunctionInfo.

Signed-off-by: Krystian Chmielewski <krystian.chmielewski@intel.com>
2022-03-07 19:05:17 +01:00
Filip Hazubski
0f54a57910 Update programming STATE_COMPUTE_MODE
Remove function clearComputeModePropertiesIfNeeded.
If a field has to be programmed unconditionally, ignore isDirty flag.

Related-To: NEO-6728

Signed-off-by: Filip Hazubski <filip.hazubski@intel.com>
2022-03-07 16:23:59 +01:00
Lukasz Jobczyk
3905fa7540 [PVC] Enable direct submission on main BCS
Signed-off-by: Lukasz Jobczyk <lukasz.jobczyk@intel.com>
2022-03-07 14:32:02 +01:00
Lukasz Jobczyk
f91ae9d59c Add multithread enqueue blit OOQ test
Signed-off-by: Lukasz Jobczyk <lukasz.jobczyk@intel.com>
2022-03-07 11:17:27 +01:00
Compute-Runtime-Validation
1a823356a3 Revert "Add prelim support for ras diagnostics and firmware"
This reverts commit 5a2145ad8d.

Signed-off-by: Compute-Runtime-Validation <compute-runtime-validation@intel.com>
2022-03-06 11:31:15 +01:00
Jaime Arteaga
a010fb3634 Dont close shared handle on imported allocations
Related-To: LOCI-2272

Signed-off-by: Jaime Arteaga <jaime.a.arteaga.molina@intel.com>
2022-03-06 04:35:47 +01:00
Bellekallu Rajkiran
7ec8d8ef91 Add prelim support for frequency events and fabric port
Related-To: LOCI-2864

Source of prelim headers:
https://github.com/intel-gpu/drm-uapi-helper/tree/master/i915-shared-headers

Signed-off-by: Bellekallu Rajkiran <bellekallu.rajkiran@intel.com>
2022-03-06 02:47:20 +01:00
Artur Harasimiuk
e07094c4c7 infra update
Related-To: NEO-6706
Signed-off-by: Artur Harasimiuk <artur.harasimiuk@intel.com>
2022-03-05 08:48:49 +01:00
Bellekallu Rajkiran
01db5ef22a Add prelim support for memory, engine and scheduler
Related-To: LOCI-2864

Source of prelim headers:
https://github.com/intel-gpu/drm-uapi-helper/tree/master/i915-shared-headers

Signed-off-by: Bellekallu Rajkiran <bellekallu.rajkiran@intel.com>
2022-03-04 19:56:30 +01:00
Michal Mrozek
a0084d4e44 Move command buffers on PVC to local memory.
Better performance of fetching commands.
Signed-off-by: Michal Mrozek <michal.mrozek@intel.com>
2022-03-04 15:26:24 +01:00
Michal Mrozek
f293c9ab25 Avoid ReadModifyWrite problem on devices with coherent L3.
Setting one of bitfields requires read from local memory which is very slow.
This is not needed for devices that have coherent L3.
Signed-off-by: Michal Mrozek <michal.mrozek@intel.com>
2022-03-04 15:05:04 +01:00